Saturday, December 26, 2009


Finite Element Model
Simulations of an electric vehicle, with varying material properties, colliding with a rigid barrier at 30 mph was performed with DYNA3D. Approximately 10,000 thin 2-D shell elements were utilized in the electric vehicle simulations as shown in Figure 1. Symmetric boundary condition was utilized about the x-y plane. Again, this was done to reduce computational time. To reduce computational time even further and since the goal of the simulations was to compare the performance of main structural members of an electric vehicle made of composite with those of steel and aluminum, components such as the engine, batteries, and seats were not modeled, but instead they were replaced by rigid blocks that would simulate the inertia of these components. An estimated components weight of 1,400 lb. was evenly distributed about the front, middle and back end of the electric vehicle. A passenger's torso was also modeled as a solid block weighing 100 lb. and placed in the driver's seat. The passenger's torso would be used to gauge the deceleration that would be incurred by a rider in the electric vehicle. This deceleration experienced by the passenger should not exceed 60 g (Standard No. 204) during the crash event [1]. The EV modeled was approximately 10 feet long, 5 feet wide and 4.5 feet tall.

Four simulations were runned. Table 1 summarizes the material properties used in the four simulations. The first simulation involved an all steel electric vehicle. The chassis and body structure, as shown in Figure 2, was modeled with structural steel properties while the other substructural members such as the bumper front fender, hood and doors were modeled with a lower grade of steel typical of a steel used for the body panels. The chassis and body structure consisted of rectangular and square cross sectioned beams. The steel chassis structure thickness was set at 0.2 cm while the body and panel thickness were respectively 0.15 cm and 0.1 cm. The next simulation modeled was an aluminum electric vehicle. Its chassis and body structure was modeled with aluminum alloy 7075 properties whereas the other substructural members were modeled with aluminum alloy 6061 properties. Its chassis thickness was set at 0.35 cm thickness and the body structure and substructure thickness were respectively 0.25 and 0.2 cm. The increase in thickness for the aluminum were done to satisfy load and deflection requirements for the electric vehicles. The third simulation involved a carbon/epoxy chassis and body structure modeled with substructures (hood, bumper, doors etc.) made of SMC composites. The composite chassis thickness was set to 0.4 cm thickness with the body structure and substructure having thickness of 0.25 and 0.2 cm. The increased thickness as compared to the steel model was again due to load and deflection requirements. The last model consisted of a carbon/epoxy chassis and body structure with glass/epoxy substructure and an SMC bumper. The thickness of the fourth model were left the same as the first composite model. For the composite models, a trigger was placed on the front rail structure, as shown in Figure 3, in order to promote stable collapse of the composite tube as demonstrated by Thornton [2]. A (0/90) laminate was utilized for the carbon/epoxy structures since this lay-up would provide the highest level of energy absorption, also reported by Thornton [2]. In addition to this, the front rail, which was 1 foot long, was also tapered in four sections. The first section near the trigger, had a thickness of 0.2 cm, in the second section the thickness was increased to 0.25 cm followed by a thickness of 0.3 cm in the third section and 0.4 cm in the last section that connects the front rail to the rest of the chassis. This was also done to promote stable collapse and has been demonstrated to be an effective method in increasing the energy absorption through dynamic sled testing of composite front rail structures [3]. It should be noted that the relative density for the composite EV structure were all above the critical value reported by Thornton [2]. Similar failure models used in the bumper study were utilized for the EV crash study as well.

Results and Discussion
Figure 4 displays the deformation on the steel EV after the crash event. A plastic hinge is formed on the front fenders of the vehicle while the bumper completely collapses. Shown in Figure 5, the front end of the chassis near the joint location is plastically bent downwards and the structure buckles near the front cross rail member. No damage is seen around the body structure. The steel EV crash behavior was as expected. The front end of a vehicle is designed to be structurally less stiff than the area around the passenger compartment so that in the event of a collision the front end serves as a sacrificial structure to dissipate the kinetic energy of the crash. As shown in Figure 5, the front chassis structure bends downward towards the main chassis structure. This bend is inherently the weak area that in the crash event will plastically bend to dissipate crash energy so that a minimum amount of this energy is transferred to the passenger compartment and the occupants.

The aluminum EV, behaves similar to the steel EV as shown in Figure 6. There is again a pronounced hinge formation in the front fender, though less pronounced, and a collapse of the bumper. The front end of the chassis (see Figure 7), does not show the dramatic buckling and hinge formation as the steel EV. In satisfying the load and displacement requirements for the EV, the aluminum EV structures thickness had to be increased. This in turn made the aluminum EV inherently stronger than the steel EV to crash loads so that less plastic deformation would be expected. The passenger compartment area is again intact. The aluminum EV with the same thickness as the steel EV did not perform well.

For the composite EV modeled with SMC substructures, as shown in Figure 8, the bumper also is collapsed inwards while the front fenders shows high buckling. A view of the chassis/body structure in Figure 9 shows that the front chassis structure as well as the passenger area entirely intact. No plastic deformation, as seen in the steel and aluminum chassis, can be seen. The stable progressive crushing, which was aimed for, can clearly be seen at the front end of the chassis as there is a high concentration of failed elements. The failed elements represent the disintegration of the composite which marks the stable collapse mode of carbon tubes. The composite EV with Glass/Epoxy substructures showed a similar behavior as seen in Figure 10 and Figure 11. The chassis/body structures is again entirely intact with no visible signs of plastic deformation.

It should be mentioned that a crash trigger mechanism at the front end of the chassis as well as a tapered front rail structure before the first cross-bar were necessary for a successful use of composite material in the EV structure.

The velocity profiles are summarized in Figures 12, 13, 14 and 15. Both composite model's velocity profile display some dips. This effect can be attributed to the tapered front rail utilized in these models. During the crash event, as the front rail crushes and reaches a new section there would be some jerk in the motion of the EV since in principle the crush event would reach a section of the rail which is slightly stronger iii than the previous section. This would also induce higher crush loads on the front rails. The steel and aluminum EVs did not display these dips since the front rail had a uniform thickness. The steel EV reached 0 m/s at about 70 milliseconds while the aluminum EV reached that at 40 milliseconds and the composite EV at around 45 milliseconds for both composite models. The lower time to reach 0 m/s, for the aluminum and composite EVs, can again be attributed to a higher strength due to a larger thickness that would enable the structure to handle higher crash loads, and this in turn leads to higher levels of deceleration of the vehicle. All models had some minimal level of residual velocity (since the velocity profiles all dip below 0) which represents the EV ricocheting off the rigid barrier.

Rigid body deceleration profiles are summarized in Figures 16, 17, 18, and 19. The maximum deceleration was experienced by the composite EV modeled with SMC substructures. A deceleration of 900 m/s2 was attained by this model. The composite EV modeled with glass/epoxy substructures reached a maximum deceleration of 750 m/s2. The aluminum EV also had a maximum deceleration of about 750 m/s2. The steel EV had the lowest maximum deceleration at 550 m/s2. The various level of deceleration experienced by the EV can again be attributed to various structural strength. Higher structural strength leads to higher crush loads to induce buckling in the steel and aluminum structure and cause crushing in the composite structures. For both the steel and aluminum EV, this maximum deceleration was attained early in the crash event whereas for both composite models this maximum deceleration occurred near the end of the crash event. This, again, can be attributed to the structural strength of the front rails due to a tapered design. As the crash phenomenon reached the front rails, in the aluminum and steel EVs, a higher crash load would be needed to start buckling, so that a higher deceleration is reached. This would show up as the sharp dip on the deceleration profile for the aluminum and steel EVs. After buckling around the front rail structure occurs, the structure would not be able to hold a high crush load so that the deceleration would be reduced as shown in the deceleration profiles (Figures 16, 17, 18, and 19). For the composite EVs, in which the strength increases along the length of the front rail, because of the tapering effect, a low crush load would be induced as the crash reached the tip of the front rail. As the crash reaches a thicker section in the front rail a higher crash load is needed to start crushing of the rails followed by a decrease in deceleration after crushing of the current section starts. This would continue until the highest crash load would be experienced at the thick end of the front rail where the strength is the highest. Therefore, for the composite EV, because of the tapering, the highest deceleration would be expected near the end of the crash event. The dips in the deceleration profile, Figures 18 and 19, corresponds to the crushing of different regions of the front rails in terms of its thickness.

The profiles of the deceleration experienced by the modeled passenger are shown in Figures 20, 21, 22 and 23. For the aluminum EV, the modeled passenger experienced the highest deceleration which was 650 m/s2 this was followed by the steel EV with a maximum deceleration of 400 m/s2. Both composite EVs had a maximum deceleration of about 260 m/s2. Maximum allowable deceleration experienced by a passenger should not exceed 60g or about 590 m/s2 [1]. All EV models satisfied this criterion with the exception of the aluminum EV.

Both composite EVs also showed relatively smooth deceleration profile. It is believed that the design of the front rail with the trigger mechanism and tapered section contributed to this phenomenon since it would produce stable progressive crushing of the tubes. However, for the steel and aluminum EV, there was a sharp dip in deceleration near the end of the crash event. This dip would correspond to the transfer of moment after the front structure of the steel and aluminum EV buckled so that the crash load is transferred onto the body structure, the stiffest area of the EV.

The kinetic energy profiles of the EVs are shown in Figures 24. All the kinetic energy profiles have the same gradual decrease of kinetic energy. Also, all of the models had some residual kinetic energy left (which is the ricocheting off the rigid barrier). From examining these profiles, it can be determined that the steel EV absorbed about 62 kJ of crash energy, the aluminum EV absorbed 55 kJ of crash energy while the composite EVs absorbed about 47 kJ of crash energy.

A structural weight comparison of the four EV models, as shown in Figure 25, has the composite EV with SMC substructures being the lightest at 86.34 kg followed by the composite EV with E-glass/epoxy substructures at 91.64 kg. The steel EV structure was the heaviest at 208.65 kg and the aluminum the second heaviest at 127.71 kg. Percentage wise, the composite EV structural weight was about 30 percent lighter than the aluminum and about 60 percent lighter than the steel structural members.

In terms of specific energy absorption, as shown in Figure 26, the composite EV structure had the highest values. About 0.57 kJ/kg was obtained for the composite with SMC substructure model while the composite with E-glass/epoxy substructure model had a value of about 0.53 kJ/kg. The steel and aluminum structures had a value of 0.30 kJ/kg and 0.43 kJ/kg, respectively.

The simulations have shown that all the models, including the composite EV, was able to absorb the energy of a 30 mph head on collision. All models were fairly intact and there was no damage to the passenger compartment areas. Also, the deceleration of the passenger model did not exceed 60 g for any of the models with the exception of the aluminum EV which had a higher passenger deceleration compared with standard values.

Based on the simulations, it has been shown that a full composite structured EV vehicle can survive a 30 mph crash without catastrophically failing. It was also shown that a trigger mechanism coupled with a tapered front rail, for the composite models, provided stable progressive crushing and this in turn kept decelerations to the passenger, in the composite model, minimal. These simulations showed that the composite structures with appropriate design can provide lighter vehicle weight, smoother deceleration and higher specific energy absorption compared with steel and aluminum structures used in EVs.

Click here to view all the EV figures.

Clicke here to view all the EV simulations.

Labels: , ,

This page is powered by Blogger. Isn't yours?

Free Counter
Free Counter